List of Tables
Chapter 3: Chapter 3: Probing the Role of Backbone Hydrogen Bonding in a Critical βsheet of the Extracellular Domain of a Cysloop Receptor
3.3 Results
III‐5
between rigidity and flexibility is believed to be central to the structure and function of allosteric proteins,8 involvement of the peptide backbone in mediating allosteric communication in the nAChR would not be surprising. Previous studies of backbone esters in ion channels have focused on the role of backbone flexibility in α‐helical domains,20‐22 seeking local movements that directly influence channel gating. Here we hope to better understand how hydrogen
bonding in a critical β‐sheet region that is distal to the channel gate influences receptor function.
Surprisingly, a number of single backbone mutations in this region were well‐tolerated, producing only modest shifts in EC50 and wild type‐like macroscopic current sizes (Imax) in the majority of cases. However, simultaneous introduction of two backbone mutations either in the same α‐strand separated by a single amino acid or at amino acids sharing a hydrogen bond between the two α‐strands led to surface‐expressed, non‐functional receptors. Together, these data suggest that while the receptor is remarkably robust in its ability to tolerate single amide‐
to‐ester mutations throughout these α‐strands, more substantial perturbations to this region have a profound effect on the protein. Therefore, it would seem that backbone movements in the outer β‐sheet are important for receptor function. This adds to a growing body of evidence that points to the importance of this region in mediating the conformational changes that emanate from the agonist binding site to the gate of the channel.
III‐6
effect on receptor function and an available α‐hydroxy analogue. The effect of the amide‐to‐
ester mutation was then compared to the analogous side chain mutant, instead of the wild‐type receptor. Most amide‐to‐ester mutations produced modest changes in EC50 (< 2‐fold) (Figure 3.3, Table 3.1). In three cases (αM144, αL199, and αV206) larger, although still not immense, shifts in the 3.5‐4‐fold range were seen. ACh‐induced whole‐cell currents of receptors with single α‐hydroxy mutations were similar to the currents produced by incorporation of the native amino acid by nonsense suppression.
Figure 3.3: Characteristics of nAChR with backbone mutations in β‐
strands 7 and 10. (A) EC50 shifts for backbone mutations made at nine positions in β7and β10. Modest effects on receptor function are seen in all cases. (B) Current traces for backbone mutations made at nine positions in β7and β10.
A
B
III‐7
EC50 (M) nH
WT 46 ± 2 1.8 ± 0.13
M144Leu 15 ± 0.3 1.7 ± 0.06
M144Lah 50 ± 0.8 1.7 ± 0.03
L146Lah 26 ± 0.8 1.3 ± 0.05
T148Tah 33 ± 3 1.0 ± 0.06
L199Lah 11 ± 0.5 1.6 ± 0.10
I201Iah 82 ± 6 1.5 ± 0.12
T202Tah 24 ± 1 1.4 ± 0.09
Y203Yah 39 ± 0.8 1.9 ± 0.07
F205Tyr 90 ± 2 1.5 ± 0.05
F205Yah 67 ±62 1.4 ± 0.13
V206Vah 170 ± 10 1.9 ± 0.20
As EC50 is a composite of both binding and gating phenomena, small shifts in EC50 may mask a large effect on receptor behavior. Recently, we developed a method to evaluate whether a loss of function (increase in EC50) mutation in the extracellular domain strongly impacts receptor gating (rather than agonist binding).23 The method, termed ELFCAR
(Elucidating Long‐range Functional Coupling in Allosteric Receptors), is based on mutant cycle analysis and produces a value for the coupling between extracellular domain residue and a known gating mutation (in this case βL9’S). This method utilizes whole‐cell data to determine whether a mutant that produces a moderate to large increase in EC50 has a substantial impact on receptor gating, as indicated by ELFCAR is especially valuable in interpreting moderate
increases in EC50, where a dramatic effect on gating might otherwise be missed. An value near one, however, does not preclude a modest effect on gating. The three α‐hydroxy mutants in this study that produce increases in EC50 – at positions 144, 201, and 206 – all produce an
value near one (0.97, 0.65, and 0.69, respectively), suggesting no dramatic impact on receptor gating. This method is not applicable to gain‐of‐function mutations, so the remaining six
Table 3.1: EC50 and Hill coefficient (±SEM) values for mutations made in this study.
III‐8
mutants could not be evaluated using ELFCAR. Combined, the data suggest that single α‐hydroxy mutations in this region do not substantially impair receptor gating.
Simultaneous incorporation of two α‐hydroxy acid mutations in α7 and α10 produces non‐
functional, surface expressed receptors
The impact of disrupting the polypeptide backbone at more than one location was investigated by simultaneously incorporating two α‐hydroxy acids. This was done in two different ways: amide‐to‐ester mutations were made either in the same α‐strand separated by a single amino acid (αM144 and αL146 in α‐strand 7), or at amino acids sharing a hydrogen bond between the two α‐strands (αL146 and αI201; Figure 3.1). Simultaneous incorporation of the same α‐hydroxy (Lah) or amino (Leu, in the case of wild‐type recovery) acid was done in both scenarios. This allowed for double suppression using the standard nonsense suppression methodology (see materials and methods). As a large shift in EC50 might have been anticipated, the double mutation experiments were carried out in the presence of the αL9’S gating mutation.
For mutations that primarily affect binding or have a only moderate effect on channel gating this mutation lowers the EC50 by about 40‐fold, allowing for reliable EC50 measurement of potentially highly deleterious mutations without complications from open channel block at high agonist concentrations.
For both double‐mutation (Lah) experiments, no whole‐cell current was observed in response to high (>1000 μM) doses of ACh. Wild‐type recovery (Leu) by nonsense suppression gave normal whole‐cell currents in the range of 5 to 15 μA. Lack of an electrophysiological response can indicate that protein folding, subunit assembly, receptor transport, or receptor function have been impacted.
III‐9
To determine whether nonfunctional mutant receptors were expressed at the plasma membrane, the localization of these mutant receptors was studied using single molecule Total Internal Reflectance Fluorescence (TIRF) microscopy. Mutant and wild type (recovery by nonsense suppression) receptors were first labelled with biotin conjugated to α‐bungarotoxin and subsequently labelled with streptavidin conjugated to the fluorophore Alexa488.
Fluorescent puncta corresponding to single receptors were clearly visible on the cell surface (Figure 3.4B). For the double α‐hydroxy‐containing receptors, the number of puncta was between 50% and 70% of the value seen for wild type receptor (recovery by nonsense
suppression) (Figure 3.4B, Table 3.2). Based on previous, extensive studies with this approach,24 it is clear that double‐mutant receptors are being surface expressed at a level such that
significant whole‐cell current would be observed, if the receptors were functional. That is, the calculated expected whole‐cell currents from the α‐hydroxy double‐mutant receptors based on puncta density are on the order of 3 to 5 μA (Table 3.2). Since we can easily detect this
magnitude of ACh‐induced current, our results clearly show that non‐functional receptors have been expressed on the oocyte surface. The calculated expected whole cell currents from the wild‐type receptors based on puncta density are on the order of 6 to 7 μA, in agreement with the electrophysiological data.
αI201L146 # Puncta/ image frame Calculated current size (μA)
Leu 265 ± 36 7.5 ± 1.0
Lah 179 ± 33 5.1 ± 0.9
76mer 51 ± 10 1.4 ± 0.3
uninjected 27 ± 5 0.8 ± 0.2
αL146M144
Leu 214 ± 15 6.1 ± 0.4
Lah 121 ± 4 3.4 ± 0.1
76mer 56 ± 5 1.6 ± 0.1
Uninjected 37 ± 4 1.0 ± 0.1
Table 3.2: Puncta densities and corresponding estimated current sizes from TIRF microscopy experiments. 76mer refers to unaminoacylated tRNA.
III‐10
Figure 3.4: Analysis of non‐functional nAChRs containing two amide‐to‐ester mutations by TIRF microscopy.
(A) Puncta density for receptors labeled with Leu refers to recovery of the wild‐type peptide backbone by nonsense suppression; Lah refers to incorporation of leucine α‐hydroxy; and 76mer indicates the negative control experiment (coinjection of mRNA with a tRNA molecule that has not been charged with an amino or hydroxy acid). The puncta density for both Leu and Lah are significantly higher than for either the 76mer or uninjected oocyte controls (t‐test values for Lah vs. 76mer are: P=0.0009 (L146I120); P<0.0001 (L146M144); for Lah vs. uninjected: P= 0.0003 (L146I120); P< 0.0001 (L146M144). Data are compiled from ≥6 non‐overlapping images from ≥2 oocytes.
(B) Representative TIRF images corresponding to the data in (A). Scale bars represent 6 μm.
B A
III‐11
Note that α‐bungarotoxin is a potent inhibitor of the nAChR that binds tightly in the same region25 as ACh and requires properly folded receptor for tight binding. Thus, successful α‐bungarotoxin binding indicates that the overall structure of the extracellular domain and binding site were not grossly perturbed. These results indicate that simultaneous
incorporation of two α‐hydroxy mutations in this region leads to nonfunctional, surface‐
expressed receptors.